\begin{tabular}{l*{4}{c}}
\hline\hline
                    &\multicolumn{1}{c}{(1)}&\multicolumn{1}{c}{(2)}&\multicolumn{1}{c}{(3)}&\multicolumn{1}{c}{(4)}\\
                    &\multicolumn{1}{c}{\% No in referendum}&\multicolumn{1}{c}{\% No in referendum}&\multicolumn{1}{c}{\% No in referendum}&\multicolumn{1}{c}{\% No in referendum}\\
\hline
M5S: referendum {it:(events only)}&        1.59&        0.52&        0.28&        0.32\\
                    &      (0.71)&      (0.22)&      (0.12)&      (0.16)\\
[1em]
M5S: \% votes 2013  &            &            &        0.25&            \\
                    &            &            &      (0.05)&            \\
[1em]
Constant            &       59.37&       59.47&       76.28&       61.41\\
                    &      (0.66)&      (0.02)&      (2.06)&      (0.10)\\
[1em]
Province FE         &            &  \checkmark&  \checkmark&  \checkmark\\
[1em]
_cons               &  \checkmark&            &            &            \\
[1em]
Controls            &            &            &  \checkmark&            \\
\hline
Obs                 &     7994.00&     7994.00&     7804.00&     7804.00\\
Provinces           &      110.00&      110.00&      110.00&      110.00\\
adj.R$^2$           &        0.00&        0.55&        0.71&        0.72\\
adj.R$^2$ (within)  &        0.00&        0.00&        0.34&        0.00\\
RMSE                &        8.73&        5.86&        4.69&        4.71\\
\hline\hline
\multicolumn{5}{l}{\footnotesize \emph{Note:} Clustered standard errors by province in parentheses. Controls omitted from table: PD: \% votes 2013, \% turnout 2013, income per cap, \% unemployed, \% university degree, \% low education, \% foreigners, population density. Same variables used for matching, history omitted from matching.}\\
\end{tabular}
